Intelligent eye-controlled bonding system and method for orthodontic appliance

Abstract

The present invention discloses an intelligent eye-controlled bonding system and method for an oral orthodontic appliance, and relates to the technical field of intelligent oral medical equipment. It comprises a pupil eye-controlled image positioning module, which generates the initial coordinates of the bracket by integrating the physician's pupil trajectory tracking with the physician's head dynamic compensation, and combines the patient's three-dimensional oral model mapping to construct a three-dimensional visualization interface by integrating the oral scanner and cone beam CT data, and projects it onto the physical tooth surface through a laser positioning projection device. The dynamic tracking module corrects the bracket posture in real time through rigid compensation and elastic compensation based on inertial sensor data. The path planning and control module adopts an improved RRT*-Connect algorithm, and integrates the elastic map and the impedance-admittance hybrid control strategy to optimize the motion path. The robotic arm execution module adopts a modular six-axis architecture, integrating a triple composite drive system and a dual-cavity end effector, and completes sub-millimeter bonding operations in the confined space of the oral cavity. The present invention improves the positioning accuracy and bonding efficiency of the bracket.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent oral medical equipment, and in particular to an intelligent eye-controlled bonding system and method for an oral orthodontic appliance. Background Art

[0002] Malocclusion is a structural abnormality of the teeth, jaws, and craniofacial structure caused by congenital genetic or environmental factors. The core of orthodontic treatment is to bond orthodontic appliances to the patient's teeth to correct malocclusion and restore normal craniofacial function. Precise positioning and bonding of the appliances are crucial in this process.

[0003] In the prior art, the methods for bonding braces mainly include direct bonding and indirect bonding. In the direct bonding method, the doctor positions and bonds the brackets directly in the patient's mouth. This method is highly dependent on the doctor's clinical experience, visual judgment, and the stability of hand operations. However, when the patient's oral environment is complex or the head moves, positioning deviations are prone to occur, resulting in reduced positioning accuracy. The indirect bonding method uses digital design and transfer trays to achieve precise positioning of the braces, but its operation process is complicated and time-consuming, thereby reducing bonding efficiency.

[0004] With the continuous advancement of computer and artificial intelligence technology, human-computer interaction methods are becoming increasingly diverse. Therefore, there is an urgent need for an intelligent eye-controlled bonding system and method for oral orthodontic appliances with both high precision and high efficiency.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0006] In response to the problems of low bracket positioning accuracy and low bonding efficiency in the existing technology, the present invention provides an intelligent eye-controlled bonding system and method for orthodontic appliances, aiming to improve the bracket positioning accuracy and bonding efficiency.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] An intelligent eye-controlled bonding system for an orthodontic appliance, comprising:

[0009] The pupil eye-controlled image positioning module is used to capture the physician's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the physician's gaze point. By monitoring the dynamic changes of the physician's head in real time, the two-dimensional coordinates of the physician's gaze point are dynamically corrected to generate the physician's gaze point compensation coordinates. The physician's gaze point compensation coordinates are mapped in real time with the patient's three-dimensional oral model to generate the initial three-dimensional coordinates of the bracket target bonding point. At the same time, the digital dental model and cone-beam CT data obtained by the intraoral scanner are integrated to construct a three-dimensional visualization interface, and the initial three-dimensional coordinates of the bracket target bonding point are projected onto the physical tooth surface in real time through a laser positioning projection device;

[0010] A dynamic tracking module is used to collect the patient's maxillofacial movement data in real time through an inertial sensor. When dynamic changes in the patient's maxillofacial region are detected, based on the initial three-dimensional coordinates of the bracket target bonding point, a corrected posture of the bracket target bonding point is generated through the superposition compensation of rigid compensation and elastic compensation, and is synchronously updated to the three-dimensional visualization interface and the laser positioning projection device; the rigid compensation calculates the displacement caused by the overall maxillofacial movement through the homogeneous coordinate transformation matrix, and corrects the initial three-dimensional coordinates of the bracket target bonding point according to the displacement to obtain the corrected three-dimensional coordinates of the bracket target bonding point; the elastic compensation is based on the biomechanical properties of the teeth, and uses finite element analysis to calculate the dentition deformation caused by the traction of the oral soft tissue, and converts the dentition deformation into a six-degree-of-freedom posture correction amount applied to the bracket target bonding point on the corrected three-dimensional coordinate, and the six-degree-of-freedom posture correction amount includes position offset and direction rotation;

[0011] A path planning and control module is used to generate an initial collision-free path between the robotic arm and the oral soft tissue based on the corrected position of the bracket target bonding point using an improved RRT*-Connect algorithm, integrate elastic map dynamic obstacle avoidance and impedance-admittance hybrid control strategy, optimize the initial collision-free path in real time, generate an optimized motion path, and display the optimized motion path on a three-dimensional visualization interface;

[0012] The robotic arm execution module is used to display the optimized motion path and the corrected position of the bracket target bonding point based on the three-dimensional visualization interface. Through a modular six-axis collaborative architecture, it integrates a triple composite drive system and a dual-cavity end effector to achieve sub-millimeter precision bracket bonding operations in the confined space of the oral cavity.

[0013] As a preferred solution of the present invention, the pupil eye control image positioning module includes: a physician visual tracking module for capturing the physician's pupil movement trajectory in real time through an infrared camera array combined with a corneal reflection algorithm, and analyzing the two-dimensional coordinates of the physician's gaze point through a deep learning model;

[0014] The head dynamic compensation module is used to monitor the dynamic changes of the doctor's head in real time, correct the two-dimensional coordinates of the doctor's gaze point through homogeneous coordinate transformation, and generate the doctor's gaze point compensation coordinates;

[0015] A coordinate mapping module is used to map the doctor's gaze point compensation coordinates to the three-dimensional coordinates on the target tooth surface through a spatial coordinate system conversion algorithm to generate the initial three-dimensional coordinates of the bracket target bonding point;

[0016] A multi-source image registration unit is used to construct a 3D visualization interface based on the initial 3D coordinates of the bracket target bonding point, integrating the digital dental model acquired by the intraoral scanner and the cone-beam CT data;

[0017] The laser positioning projection module is used to project the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time through a laser positioning projection device.

[0018] As a preferred solution of the present invention, the improved RRT*-Connect algorithm includes:

[0019] Building rapidly expanding random trees and ; The tree is constructed with the current pose of the robot end as the root node; The tree is constructed with the corrected pose of the bracket target bonding point as the root node;

[0020] Calculate the stenosis coefficient of the current expansion direction , dynamically adjust the expansion step size, the formula is:

[0021]

[0022] Where, The distance between the robotic arm and the nearest oral soft tissue in the current extension direction; is the range of perception; is the minimum value of the expansion step; is the maximum value of the expansion step; is the dynamically adjusted expansion step size;

[0023] Sampling is performed in a 90-degree cone area in the target direction with a probability of 0.7, and random sampling is performed in the entire space with a probability of 0.3;

[0024] when and When the distance between the nearest nodes is less than the preset threshold, the connection and Generate an initial collision-free path between the robotic arm and oral soft tissue.

[0025] As a preferred solution of the present invention, elastic map dynamic obstacle avoidance includes:

[0026] The minimum safe distance between the robotic arm and oral soft tissue is calculated in real time using the formula:

[0027]

[0028] Where, For the current moment The minimum safe distance between the robotic arm and oral soft tissue, is the three-dimensional coordinate of the tip of the robotic arm, For the The coordinates of the oral soft tissue sampling points, It is a collection of all oral soft tissue sampling points;

[0029] when Less than the preset safety distance threshold When , the path point to be optimized is selected from the initial collision-free path , and generate the optimized path point through the attraction of the bracket target bonding point and the repulsion of the oral soft tissue sampling point. The formula is:

[0030] ;

[0031] Where, is the optimized path point; is the coordinate of the path point to be optimized; is the coordinate of the target bonding point; is the weight coefficient of the attraction of the target bonding point; is the weight coefficient of the repulsive force at the soft tissue sampling point; is the number of soft tissue sampling points; is a very small constant;

[0032] Will Alternative Update the initial collision-free path to generate an optimized motion path.

[0033] As a preferred solution of the present invention, the impedance-admittance hybrid control strategy includes:

[0034] The contact force between the robotic arm and the target bonding point of the bracket on the tooth surface is monitored in real time through a six-dimensional force sensor. When the vertical pressure exceeds 0.5N or the lateral shear force exceeds 0.3N, it switches to the compliant control mode within 5ms.

[0035] The angular velocity of the robot joint is calculated by the Jacobian matrix and the pseudo-inverse of the Jacobian matrix. The formula is:

[0036]

[0037] Where, is the joint angular velocity vector of the six-axis robot arm; is the Jacobian matrix; is the pseudo-inverse of the Jacobian matrix, , is the transposed matrix of the Jacobian matrix, is the damping coefficient; is the desired velocity vector of the tip of the robotic arm; is the weight coefficient; is the secondary task vector, used to optimize the joint configuration; is the identity matrix;

[0038] The deviation between the position of the tip of the constrained robot arm and the target bonding point of the bracket is calculated as follows:

[0039]

[0040] Where, The tip of the robotic arm at time The three-dimensional coordinates of is the corrected three-dimensional coordinate of the bracket target bonding point, extracted from the corrected pose of the bracket target bonding point; is the maximum allowable deviation; is the starting time point; The end time point.

[0041] As a preferred embodiment of the present invention, the triple composite drive system includes:

[0042] Piezoelectric ceramic micro-motion stage for Z-axis feed, and bonding pressure is controlled by laser interferometer closed-loop feedback;

[0043] Magnetorheological flexible joints for achieving Dynamic inclination compensation, real-time adaptation to changes in tooth surface curvature;

[0044] Rotating ultrasonic motor is used to achieve 0.001° level attitude fine adjustment.

[0045] As a preferred embodiment of the present invention, the dual-cavity end effector includes:

[0046] The upper cavity has a built-in vacuum adsorption array for grabbing brackets of different specifications;

[0047] The lower cavity integrates a microfluidic coating module, which controls the adhesive dosage through a micro-electromechanical system micro pump to achieve dot-matrix coating on the tooth surface.

[0048] As a preferred solution of the present invention, the system further includes a bonding quality closed-loop control unit:

[0049] The contact stress distribution between the bracket and the tooth surface is monitored in real time by using a fiber Bragg grating sensor;

[0050] The adhesive flow state is detected by confocal microscopy, and automatic glue filling is triggered when bubbles or defects are found;

[0051] A three-step pressure curve was implemented to achieve pressure curing; the three-step pressure curve included an initial contact period with a pre-pressure of 0.1N, an adhesive wetting period with a linear pressure increase to 0.5N, and a final curing period with a pressure maintained at 1.2N.

[0052] As a preferred solution of the present invention, the system also includes a biosafety protection unit for covering the surface of the robotic arm with a medical-grade silicone isolation film, and integrating an ultraviolet LED (Light Emitting Diode) dynamic sterilization module at the end.

[0053] A bonding method of an intelligent eye-controlled bonding system for an orthodontic appliance, comprising:

[0054] S1 uses an infrared camera array combined with a corneal reflection algorithm to capture the doctor's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the doctor's gaze point through a deep learning model;

[0055] S2, real-time monitoring of the dynamic changes of the physician's head, and correction of the two-dimensional coordinates of the physician's gaze point through homogeneous coordinate transformation to generate the compensated coordinates of the physician's gaze point;

[0056] S3, through the spatial coordinate system conversion algorithm, the doctor's gaze point compensation coordinates are mapped to the three-dimensional coordinates on the target tooth surface to generate the initial three-dimensional coordinates of the bracket target bonding point;

[0057] S4, based on the initial three-dimensional coordinates of the bracket target bonding point, the digital dental model obtained by the intraoral scanner and the cone-beam CT data were integrated to construct a three-dimensional visualization interface;

[0058] S5, projecting the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time using a laser positioning projection device;

[0059] S6, dynamically tracking the patient's maxillofacial movements. When dynamic changes in the patient's maxillofacial region are detected, the corrected position of the bracket's target bonding point is generated based on the initial three-dimensional coordinates of the bracket's target bonding point through the superposition of rigid compensation and elastic compensation, and is synchronously updated to the three-dimensional visualization interface and laser positioning projection device;

[0060] S7, based on the corrected position of the bracket target bonding point, using the improved RRT*-Connect algorithm to generate an initial collision-free path between the robotic arm and the oral soft tissue, integrating the elastic map dynamic obstacle avoidance and the impedance-admittance hybrid control strategy, optimizing the initial collision-free path in real time, generating an optimized motion path, and displaying the optimized motion path on a three-dimensional visualization interface;

[0061] The S8 adopts a modular six-axis collaborative architecture, integrates a triple composite drive system and a dual-cavity end effector, and performs bracket bonding operations with sub-millimeter precision in the confined space of the oral cavity based on the optimized motion path and the corrected position of the bracket target bonding point displayed on a three-dimensional visualization interface.

[0062] Compared with the existing technology, the beneficial effects of the present invention are: by capturing the doctor's pupil movement trajectory in real time, integrating the doctor's head dynamic compensation, and combining the patient's three-dimensional oral model mapping technology, the doctor's clinical experience is converted into sub-millimeter spatial coordinates, effectively eliminating the visual positioning deviation and improving the accuracy of the initial positioning of the bracket; using inertial sensors to monitor the patient's maxillofacial movement, and combining the dual mechanisms of rigid compensation and elastic compensation, it effectively overcomes the interference of the patient's head shaking and oral soft tissue deformation on positioning, and improves the adaptability and positioning accuracy of the bracket in a dynamic environment; using the improved RRT*-Connect algorithm to generate the initial collision-free path between the robotic arm and the oral soft tissue, integrating the elastic map Dynamic obstacle avoidance technology can sense changes in the oral environment in real time, optimize the motion path between the robotic arm and oral soft tissue, and combine with the impedance-admittance hybrid control strategy to achieve adaptive contact between the robotic arm and oral soft tissue, reducing the risk of damage to oral soft tissue by the robotic arm; a modular six-axis collaborative robotic arm is used, integrating a triple composite drive system to achieve nanometer-level displacement compensation, ensuring sub-millimeter-level precise positioning and operation in a restricted and dynamic oral environment; the dual-cavity end effector, with the assistance of a three-dimensional visualization interface, combined with a laser positioning projection device, simultaneously completes the quantitative injection of adhesive and the adsorption positioning of the bracket. The integrated operation process effectively shortens the bracket bonding time and improves the bracket bonding efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a structural diagram of an intelligent eye-controlled bonding system for orthodontic appliances proposed by the present invention;

[0064] Figure 2 This is a flow chart of the dynamic tracking compensation mechanism proposed by the present invention;

[0065] Figure 3 This is a flow chart of the initial collision-free path planning algorithm proposed by the present invention;

[0066] Figure 4 This is a flow chart of the bonding quality control method proposed by the present invention;

[0067] Figure 5 This is a flow chart of a bonding method of an intelligent eye-controlled bonding system for an orthodontic appliance proposed by the present invention. DETAILED DESCRIPTION

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of the present invention.

[0069] In the prior art, direct bonding involves the doctor positioning and bonding the brackets directly in the patient's mouth. However, positioning errors can occur when the patient's oral environment is complex or the patient's head moves, thereby reducing the positioning accuracy of the appliance. The indirect bonding method achieves precise positioning of the appliance through digital design and transfer trays, but its operation process is complex and time-consuming, resulting in reduced bonding efficiency. The present invention proposes an intelligent eye-controlled bonding system for orthodontic appliances. The system uses an infrared camera and corneal reflection algorithm to capture the physician's pupil movement trajectory in real time, combines a deep learning model to analyze the physician's gaze point, integrates dynamic compensation for the physician's head, and combines the patient's three-dimensional oral model mapping technology to generate accurate initial three-dimensional coordinates of the bracket target bonding point. Inertial sensors are used to monitor maxillofacial movement, and a dual mechanism of rigid compensation and elastic compensation is combined to effectively cope with the patient's head movement, improving adaptability to dynamic environments and the accuracy of bracket positioning. An improved RRT*-Connect algorithm and elastic map technology are used, combined with an impedance-admittance control strategy, to optimize the motion path between the robotic arm and oral soft tissue, thereby reducing damage to the oral soft tissue caused by the robotic arm. The modular six-axis robotic arm and triple-composite drive system achieve nanometer-level displacement compensation, further improving bracket positioning accuracy. The dual-cavity end effector simultaneously performs adhesive injection and bracket adsorption positioning, improving bracket bonding efficiency.

[0070] like Figure 1 As shown in FIG, an embodiment of the present invention provides an intelligent eye-controlled bonding system for an orthodontic appliance, including a pupil eye-controlled image positioning module, a dynamic tracking module, a path planning and control module, and a robotic arm execution module.

[0071] (1) Pupil eye control image positioning module

[0072] It is used to capture the doctor's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the doctor's gaze point. By monitoring the dynamic changes of the doctor's head in real time, it dynamically corrects the two-dimensional coordinates of the doctor's gaze point, generates the doctor's gaze point compensation coordinates, and maps the doctor's gaze point compensation coordinates with the patient's three-dimensional oral model in real time to generate the initial three-dimensional coordinates of the bracket target bonding point. At the same time, it integrates the digital dental model and cone beam CT data obtained by the intraoral scanner to construct a three-dimensional visualization interface, and projects the initial three-dimensional coordinates of the bracket target bonding point to the physical tooth surface in real time through the laser positioning projection device.

[0073] The pupil eye control image positioning module includes a physician vision tracking module, a head dynamic compensation module, a coordinate mapping module and a multi-source image registration unit.

[0074] The physician vision tracking module uses an infrared camera array combined with a corneal reflection algorithm to capture the physician's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the physician's gaze point using a deep learning model;

[0075] The head dynamic compensation module is used to monitor the dynamic changes of the doctor's head in real time, correct the two-dimensional coordinates of the doctor's gaze point through homogeneous coordinate transformation, and generate the doctor's gaze point compensation coordinates;

[0076] A coordinate mapping module is used to map the doctor's gaze point compensation coordinates to the three-dimensional coordinates on the target tooth surface through a spatial coordinate system conversion algorithm to generate the initial three-dimensional coordinates of the bracket target bonding point;

[0077] A multi-source image registration unit is used to construct a 3D visualization interface based on the initial 3D coordinates of the bracket target bonding point, integrating the digital dental model acquired by the intraoral scanner and the cone-beam CT data;

[0078] The laser positioning projection module is used to project the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time through a laser positioning projection device.

[0079] Furthermore, this module, located at the front end of the robotic arm, uses a laser positioning projection device to project the initial three-dimensional coordinates of the bracket's target bonding point onto the physical tooth surface in real time, assisting the physician in quickly determining the initial position of the bracket's target bonding point. Furthermore, this module includes a built-in standardized operation database, supporting parameter presets for different bracket systems.

[0080] (2) Dynamic tracking module

[0081] It is used to collect the patient's maxillofacial movement data in real time through inertial sensors. When dynamic changes in the patient's maxillofacial region are detected, based on the initial three-dimensional coordinates of the bracket target bonding point, the corrected posture of the bracket target bonding point is generated through the superposition compensation of rigid compensation and elastic compensation, and synchronously updated to the three-dimensional visualization interface and laser positioning projection device; the rigid compensation calculates the displacement caused by the overall movement of the maxillofacial region through the homogeneous coordinate transformation matrix, and corrects the initial three-dimensional coordinates of the bracket target bonding point according to the displacement to obtain the corrected three-dimensional coordinates of the bracket target bonding point; the elastic compensation is based on the biomechanical properties of the teeth, and uses finite element analysis to calculate the dentition deformation caused by the traction of the oral soft tissue, and converts the dentition deformation into a six-degree-of-freedom posture correction amount applied to the bracket target bonding point on the corrected three-dimensional coordinate, and the six-degree-of-freedom posture correction amount includes position offset and direction rotation.

[0082] Furthermore, if Figure 2 As shown in the figure, when the dynamic changes of the patient's maxillofacial region are detected, the corrected position of the bracket target bonding point is generated based on the initial three-dimensional coordinates of the bracket target bonding point through the superposition compensation of rigid compensation and elastic compensation. The specific implementation process is as follows:

[0083] The displacement caused by the overall movement of the maxillofacial region is calculated by the homogeneous coordinate transformation matrix, and the initial three-dimensional coordinates of the bracket target bonding point are corrected according to the displacement to obtain the corrected three-dimensional coordinates of the bracket target bonding point. The formula is:

[0084] ;

[0085] ;

[0086] Where, is the homogeneous coordinate transformation matrix; is the rotation matrix, is the translation vector; is the initial three-dimensional coordinate of the bracket target bonding point; The corrected three-dimensional coordinates of the bracket target bonding point after rigid compensation;

[0087] Calculation of tooth surface deformation vectors using dynamic equations , the kinetic equation is:

[0088] ;

[0089] Where, is the mass matrix; is the damping matrix; is the stiffness matrix; is the deformation vector; for The second derivative of for The first derivative of ; is the soft tissue tension force that changes with time;

[0090] The deformation vector Converted into a six-degree-of-freedom pose correction applied to the bracket target bonding point on the corrected three-dimensional coordinates , including position offset and direction rotation, the formula is:

[0091] ;

[0092] Where, is the transformation matrix;

[0093] Calculate the corrected position of the bracket's target bonding point , the formula is:

[0094] .

[0095] This module is located at the front end of the robotic arm. When the patient's head moves unexpectedly, such as a slight shake, rigid compensation and elastic compensation are superimposed for compensation, and the compensation delay is less than 10ms.

[0096] (3) Path planning and control module

[0097] It is used to generate an initial collision-free path between the robotic arm and oral soft tissue based on the corrected posture of the bracket target bonding point, using an improved RRT*-Connect algorithm, integrating elastic map dynamic obstacle avoidance and impedance-admittance hybrid control strategy, optimizing the initial collision-free path in real time, generating an optimized motion path, and displaying the optimized motion path on a three-dimensional visualization interface.

[0098] like Figure 3 As shown in Figure 2, the improved RRT*-Connect algorithm includes:

[0099] Building rapidly expanding random trees and ; The tree is constructed with the current pose of the robot end as the root node; The tree is constructed with the corrected pose of the bracket target bonding point as the root node;

[0100] Calculate the stenosis coefficient of the current expansion direction , dynamically adjust the expansion step size, the formula is:

[0101] ;

[0102] Where, The distance between the robotic arm and the nearest oral soft tissue in the current extension direction; is the range of perception; is the minimum value of the expansion step; is the maximum value of the expansion step; is the dynamically adjusted expansion step size;

[0103] Sampling is performed in a 90-degree cone area in the target direction with a probability of 0.7, and random sampling is performed in the entire space with a probability of 0.3;

[0104] when and When the distance between the nearest nodes is less than the preset threshold, the connection and Generate an initial collision-free path between the robotic arm and oral soft tissue.

[0105] Elastic map dynamic obstacle avoidance includes:

[0106] The minimum safe distance between the robotic arm and oral soft tissue is calculated in real time using the formula:

[0107] ;

[0108] Where, For the current moment The minimum safe distance between the robotic arm and oral soft tissue, is the three-dimensional coordinate of the tip of the robotic arm, For the The coordinates of the oral soft tissue sampling points, It is a collection of all oral soft tissue sampling points;

[0109] when Less than the preset safety distance threshold When , the path point to be optimized is selected from the initial collision-free path , and generate the optimized path point through the attraction of the bracket target bonding point and the repulsion of the oral soft tissue sampling point. The formula is:

[0110] ;

[0111] Where, is the optimized path point; is the coordinate of the path point to be optimized; is the coordinate of the target bonding point; is the weight coefficient of the attraction of the target bonding point; is the weight coefficient of the repulsive force at the soft tissue sampling point; is the number of soft tissue sampling points; is a very small constant;

[0112] Will Alternative Update the initial collision-free path to generate an optimized motion path.

[0113] Furthermore, the safety distance preset threshold The setting needs to be based on the specific environment of the patient's oral cavity and the size of the instrument. The recommended value is 5mm.

[0114] The impedance-admittance hybrid control strategy includes:

[0115] The contact force between the robotic arm and the target bonding point of the bracket on the tooth surface is monitored in real time through a six-dimensional force sensor. When the vertical pressure exceeds 0.5N or the lateral shear force exceeds 0.3N, it switches to the compliant control mode within 5ms.

[0116] The angular velocity of the robot joint is calculated by the Jacobian matrix and the pseudo-inverse of the Jacobian matrix. The formula is:

[0117] ;

[0118] Where, is the joint angular velocity vector of the six-axis robot arm; is the Jacobian matrix; is the pseudo-inverse of the Jacobian matrix, , is the transposed matrix of the Jacobian matrix, is the damping coefficient; is the desired velocity vector of the tip of the robotic arm; is the weight coefficient; is the secondary task vector, used to optimize the joint configuration; is the identity matrix;

[0119] The deviation between the position of the tip of the constrained robot arm and the target bonding point of the bracket is calculated as follows:

[0120] ;

[0121] Where, The tip of the robotic arm at time The three-dimensional coordinates of is the corrected three-dimensional coordinate of the bracket target bonding point, extracted from the corrected pose of the bracket target bonding point; is the maximum allowable deviation; is the starting time point; The end time point.

[0122] Furthermore, Usually less than 0.5mm.

[0123] This module, located at the front end of the robotic arm, monitors the dynamic deformation of oral soft tissues during motion trajectory execution through a fusion of binocular vision and millimeter-wave radar. A pharyngeal-cheek motion prediction model, built using deep learning technology, can predict the displacement trends of vulnerable areas such as the tongue and buccal mucosa 200 milliseconds in advance. When the safe distance between the robotic arm and oral soft tissue drops below the 5mm threshold, the module dynamically adjusts the motion trajectory based on an elastic map algorithm, replanning the obstacle avoidance path with submillimeter accuracy. To address the strong nonlinear characteristics of the oral operating space, the module incorporates a built-in impedance-admittance hybrid control strategy. Using a six-axis force sensor, the module monitors the contact force between the robotic arm and the bracket target bonding point on the tooth surface in real time. If vertical pressure exceeds 0.5N or lateral shear force exceeds 0.3N, the module adaptively adjusts the robotic arm's stiffness parameters and switches to compliant control mode within 5ms to prevent damage to the tooth enamel or mucosa caused by sudden patient movements. Simultaneously, an inverse kinematics solver synchronously updates the angular velocity of each joint of the six-axis robotic arm, ensuring that the robotic arm tip remains precisely locked to the bracket target bonding point during posture adjustments.

[0124] (4) Robotic arm execution module

[0125] It is used to display the optimized motion path and the corrected position of the bracket target bonding point based on the three-dimensional visualization interface. Through the modular six-axis collaborative architecture, it integrates the triple composite drive system and the dual-cavity end effector to achieve sub-millimeter precision bracket bonding operations in the confined space of the oral cavity.

[0126] The triple composite drive system includes a piezoelectric ceramic micro-motion platform, a magnetorheological flexible joint and a rotary ultrasonic motor.

[0127] The piezoelectric ceramic micro-motion stage is used for Z-axis feed, and the bonding pressure is controlled by laser interferometer closed-loop feedback.

[0128] Magnetorheological flexible joints for achieving Dynamic inclination compensation, real-time adaptation to changes in tooth surface curvature.

[0129] Rotating ultrasonic motor is used to achieve 0.001° level attitude fine adjustment.

[0130] The dual-cavity end effector includes an upper cavity and a lower cavity.

[0131] The upper cavity has a built-in vacuum adsorption array for grabbing brackets of different specifications.

[0132] The lower cavity integrates a microfluidic coating module, which controls the adhesive dosage through a micro-electromechanical system micro pump to achieve dot-matrix coating on the tooth surface.

[0133] Furthermore, the upper cavity is equipped with a built-in vacuum adsorption array with an adsorption force between 0.5 and 2N, which can accurately grasp brackets of different specifications; the lower cavity is integrated with a microfluidic coating module, which accurately controls the adhesive dosage through a micro-electromechanical system micropump with a flow accuracy of ±0.1μL. At the same time, it cooperates with a 316L stainless steel microneedle array to achieve dot matrix coating on the tooth surface, ensuring that the adhesive diffusion area accurately matches the bracket base morphology.

[0134] (5) Adhesion quality closed-loop control unit

[0135] The contact stress distribution between the bracket and the tooth surface is monitored in real time by using a fiber Bragg grating sensor;

[0136] The adhesive flow state is detected by confocal microscopy, and automatic glue filling is triggered when bubbles or defects are found;

[0137] A three-step pressure curve was applied to achieve pressure curing, which included an initial contact period with a pre-pressure of 0.1 N, an adhesive wetting period with a linear pressure increase to 0.5 N, and a final curing period with a pressure maintained at 1.2 N.

[0138] like Figure 4As shown, during the bonding process, the robotic arm monitors the contact stress distribution between the bracket and the tooth surface in real time through a fiber Bragg grating strain sensor. When it detects that the local pressure deviation exceeds the preset threshold, the magnetorheological flexible joint in the triple composite drive system immediately activates the impedance adaptive algorithm, adjusting the stiffness coefficient of each joint within 5ms to eliminate stress concentration caused by tooth surface morphological variations. The setting of the preset threshold needs to be calibrated according to the tooth position and adhesive. Its typical value range is 30% to 50%, and 40% is often used as the benchmark value. At the same time, the high-frame rate confocal microscope integrated at the end, with a scanning rate of 120fps, continuously monitors the bonding interface, analyzes the adhesive flow state through a deep learning model, and automatically triggers the glue filling program when bubbles or edge defects are found, ensuring that the uniformity of the bonding layer reaches more than 98%. Finally, the third-order pressure curve is automatically executed during the pressing stage: during the initial contact period from 0 seconds to 0.5 seconds, a pre-pressure of 0.1 Newton is used to ensure the positioning of the bracket; during the adhesive infiltration period from 0.5 seconds to 3 seconds, the pressure is linearly increased to 0.5 Newton to promote resin flow; during the final curing period from 3 seconds to 5 seconds, a constant pressure of 1.2 Newton is maintained to ensure bonding strength. The pressure fluctuation during the whole process is controlled within ±0.05 Newton.

[0139] (6) Biosafety Protection Unit

[0140] Used to cover the surface of the robotic arm with a medical-grade silicone isolation film, and integrate a UV LED dynamic sterilization module at the end.

[0141] Furthermore, to meet the biosafety requirements of the oral environment, the actuator adopts a fully enclosed sterile protection design: the surface of the robotic arm is covered with a medical-grade silicone isolation film; the end is equipped with an ultraviolet LED dynamic sterilization module with a wavelength of 265nm, which continuously kills pathogenic microorganisms at an irradiation intensity of 10mW per square centimeter during the non-contact stage.

[0142] like Figure 5 FIG. 1 is another embodiment of the present invention, which provides a bonding method of an intelligent eye-controlled bonding system for an orthodontic appliance, comprising:

[0143] S1 uses an infrared camera array combined with a corneal reflection algorithm to capture the doctor's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the doctor's gaze point through a deep learning model;

[0144] S2, real-time monitoring of the dynamic changes of the physician's head, and correction of the two-dimensional coordinates of the physician's gaze point through homogeneous coordinate transformation to generate the compensated coordinates of the physician's gaze point;

[0145] S3, through the spatial coordinate system conversion algorithm, the doctor's gaze point compensation coordinates are mapped to the three-dimensional coordinates on the target tooth surface to generate the initial three-dimensional coordinates of the bracket target bonding point;

[0146] S4, based on the initial three-dimensional coordinates of the bracket target bonding point, the digital dental model obtained by the intraoral scanner and the cone-beam CT data were integrated to construct a three-dimensional visualization interface;

[0147] S5, projecting the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time using a laser positioning projection device;

[0148] S6, dynamically tracking the patient's maxillofacial movements. When dynamic changes in the patient's maxillofacial region are detected, the corrected position of the bracket's target bonding point is generated based on the initial three-dimensional coordinates of the bracket's target bonding point through the superposition of rigid compensation and elastic compensation, and is synchronously updated to the three-dimensional visualization interface and laser positioning projection device;

[0149] S7, based on the corrected position of the bracket target bonding point, an improved RRT*-Connect algorithm is used to generate an initial collision-free path between the robotic arm and the oral soft tissue, and the elastic map dynamic obstacle avoidance and impedance-admittance hybrid control strategy are integrated to optimize the initial collision-free path in real time, generate an optimized motion path, and display the optimized motion path on a three-dimensional visualization interface;

[0150] The S8 adopts a modular six-axis collaborative architecture, integrates a triple composite drive system and a dual-cavity end effector, and performs bracket bonding operations with sub-millimeter precision in the confined space of the oral cavity based on the optimized motion path and the corrected position of the bracket target bonding point displayed on a three-dimensional visualization interface.

[0151] In summary, the doctor's pupil movement trajectory is captured in real time by an infrared camera and corneal reflection algorithm, the doctor's gaze point is analyzed by combining a deep learning model, and the doctor's head dynamic compensation is integrated. Combined with the patient's three-dimensional oral model mapping technology, the doctor's clinical experience is converted into sub-millimeter spatial coordinates, effectively eliminating visual positioning deviation and improving the initial positioning accuracy of the bracket target bonding point; inertial sensors are used to monitor maxillofacial movement. When changes in the maxillofacial region are detected, the dual mechanisms of rigid compensation and elastic compensation are combined to effectively overcome the interference of the patient's head shaking and oral soft tissue deformation on positioning, thereby improving the adaptability to dynamic environments and the positioning accuracy of the bracket target bonding point; the improved RRT* is used -Connect algorithm, integrated with elastic map dynamic obstacle avoidance technology, optimizes the motion path between the robotic arm and oral soft tissue in real time, and combines it with an impedance-admittance hybrid control strategy to achieve adaptive contact between the robotic arm and oral soft tissue, reducing damage to oral soft tissue caused by the robotic arm; a modular six-axis collaborative robotic arm is used, integrating a triple composite drive system to achieve nanometer-level displacement compensation, and realize submillimeter-level operation in a restricted and dynamic oral environment; the dual-cavity end effector, with the assistance of a three-dimensional visualization interface and combined with a laser positioning projection device, simultaneously completes the quantitative injection of adhesive and the adsorption positioning of the bracket. The integrated operation process effectively shortens the bracket bonding time and improves the bonding efficiency.

[0152] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware, or any other combination. When implemented using software, all or part of the embodiments may be implemented in the form of a computer program product, which includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to the present application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0153] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the aforementioned integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium. The storage medium may be a read-only memory, a magnetic disk, or an optical disk, etc.

[0154] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included within the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An intelligent eye-controlled bonding system for an orthodontic appliance, comprising a robotic arm, characterized in that: include: The pupil eye-controlled image positioning module is used to capture the physician's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the physician's gaze point. By monitoring the dynamic changes of the physician's head in real time, the two-dimensional coordinates of the physician's gaze point are dynamically corrected to generate the physician's gaze point compensation coordinates. The physician's gaze point compensation coordinates are mapped in real time with the patient's three-dimensional oral model to generate the initial three-dimensional coordinates of the bracket target bonding point. At the same time, the digital dental model and cone-beam CT data obtained by the intraoral scanner are integrated to construct a three-dimensional visualization interface, and the initial three-dimensional coordinates of the bracket target bonding point are projected onto the physical tooth surface in real time through a laser positioning projection device; The dynamic tracking module is used to collect the patient's maxillofacial movement data in real time through inertial sensors. When dynamic changes in the patient's maxillofacial area are detected, the module generates a corrected position of the bracket's target bonding point based on the initial three-dimensional coordinates of the bracket's target bonding point through the superposition of rigid compensation and elastic compensation, and synchronously updates it to the three-dimensional visualization interface and laser positioning projection device; The rigid compensation calculates the displacement caused by the overall movement of the maxillofacial area through a homogeneous coordinate transformation matrix, and corrects the initial three-dimensional coordinates of the bracket target bonding point according to the displacement to obtain the corrected three-dimensional coordinates of the bracket target bonding point; the elastic compensation is based on the biomechanical properties of the teeth, and uses finite element analysis to calculate the dentition deformation caused by the traction of the oral soft tissue, and converts the dentition deformation into a six-degree-of-freedom posture correction amount applied to the bracket target bonding point on the corrected three-dimensional coordinate, and the six-degree-of-freedom posture correction amount includes a position offset and a direction rotation; A path planning and control module is used to generate an initial collision-free path between the robotic arm and the oral soft tissue based on the corrected position of the bracket target bonding point using an improved RRT*-Connect algorithm, integrate elastic map dynamic obstacle avoidance and impedance-admittance hybrid control strategy, optimize the initial collision-free path in real time, generate an optimized motion path, and display the optimized motion path on a three-dimensional visualization interface; The robotic arm execution module is used to display the optimized motion path and the corrected position of the bracket target bonding point based on the three-dimensional visualization interface. Through a modular six-axis collaborative architecture, it integrates a triple composite drive system and a dual-cavity end effector to achieve sub-millimeter precision bracket bonding operations in the confined space of the oral cavity.

2. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 1, characterized in that: The pupil eye control image positioning module includes: The physician vision tracking module uses an infrared camera array combined with a corneal reflection algorithm to capture the physician's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the physician's gaze point using a deep learning model; The head dynamic compensation module is used to monitor the dynamic changes of the doctor's head in real time, correct the two-dimensional coordinates of the doctor's gaze point through homogeneous coordinate transformation, and generate the doctor's gaze point compensation coordinates; A coordinate mapping module is used to map the doctor's gaze point compensation coordinates to the three-dimensional coordinates on the target tooth surface through a spatial coordinate system conversion algorithm to generate the initial three-dimensional coordinates of the bracket target bonding point; A multi-source image registration unit is used to construct a 3D visualization interface based on the initial 3D coordinates of the bracket target bonding point, integrating the digital dental model acquired by the intraoral scanner and the cone-beam CT data; The laser positioning projection module is used to project the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time through a laser positioning projection device.

3. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 1, characterized in that: The improved RRT*-Connect algorithm includes: Building rapidly expanding random trees and ; The tree is constructed with the current pose of the robot end as the root node; The tree is constructed with the corrected pose of the bracket target bonding point as the root node; Calculate the stenosis coefficient of the current expansion direction , dynamically adjust the expansion step size, the formula is: ； Where, The distance between the robotic arm and the nearest oral soft tissue in the current extension direction; is the range of perception; is the minimum value of the expansion step; is the maximum value of the expansion step; is the dynamically adjusted expansion step size; Sampling is performed in a 90-degree cone area in the target direction with a probability of 0.7, and random sampling is performed in the entire space with a probability of 0.3; when and When the distance between the nearest nodes is less than the preset threshold, the connection and Generate an initial collision-free path between the robotic arm and oral soft tissue.

4. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 3, characterized in that: The elastic map dynamic obstacle avoidance includes: The minimum safe distance between the robotic arm and oral soft tissue is calculated in real time using the formula: ； Where, For the current moment The minimum safe distance between the robotic arm and oral soft tissue, is the three-dimensional coordinate of the tip of the robotic arm, For the The coordinates of the oral soft tissue sampling points, It is a collection of all oral soft tissue sampling points; when Less than the preset safety distance threshold When , the path point to be optimized is selected from the initial collision-free path , and generate the optimized path point through the attraction of the bracket target bonding point and the repulsion of the oral soft tissue sampling point. The formula is: ； Where, is the optimized path point; is the coordinate of the path point to be optimized; is the coordinate of the target bonding point; is the weight coefficient of the attraction of the target bonding point; is the weight coefficient of the repulsive force at the soft tissue sampling point; is the number of soft tissue sampling points; is a very small constant; Will Alternative Update the initial collision-free path to generate an optimized motion path.

5. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 4, characterized in that: The impedance-admittance hybrid control strategy includes: The contact force between the robotic arm and the target bonding point of the bracket on the tooth surface is monitored in real time through a six-dimensional force sensor. When the vertical pressure exceeds 0.5N or the lateral shear force exceeds 0.3N, it switches to the compliant control mode within 5ms. The angular velocity of the robot joint is calculated by the Jacobian matrix and the pseudo-inverse of the Jacobian matrix. The formula is: ； Where, is the joint angular velocity vector of the six-axis robot arm; is the Jacobian matrix; is the pseudo-inverse of the Jacobian matrix, , is the transposed matrix of the Jacobian matrix, is the damping coefficient; is the desired velocity vector of the tip of the robotic arm; is the weight coefficient; is the secondary task vector, used to optimize the joint configuration; is the identity matrix; The deviation between the position of the tip of the constrained robot arm and the target bonding point of the bracket is calculated as follows: ； Where, The tip of the robotic arm at time The three-dimensional coordinates of is the corrected three-dimensional coordinate of the bracket target bonding point, extracted from the corrected pose of the bracket target bonding point; is the maximum allowable deviation; is the starting time point; The end time point.

6. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 1, characterized in that: The triple composite drive system comprises: Piezoelectric ceramic micro-motion stage for Z-axis feed, and bonding pressure is controlled by laser interferometer closed-loop feedback; Magnetorheological flexible joints for achieving Dynamic inclination compensation, real-time adaptation to changes in tooth surface curvature; Rotating ultrasonic motor is used to achieve 0.001° level attitude fine adjustment.

7. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 6, characterized in that: The dual-cavity end effector comprises: The upper cavity has a built-in vacuum adsorption array for grabbing brackets of different specifications; The lower cavity integrates a microfluidic coating module, which controls the adhesive dosage through a micro-electromechanical system micro pump to achieve dot-matrix coating on the tooth surface.

8. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 1, characterized in that: The system also includes a bonding quality closed-loop control unit: The contact stress distribution between the bracket and the tooth surface is monitored in real time by using a fiber Bragg grating sensor; The adhesive flow state is detected by confocal microscopy, and automatic glue filling is triggered when bubbles or defects are found; A three-step pressure curve was implemented to achieve pressure curing; the three-step pressure curve included an initial contact period with a pre-pressure of 0.1N, an adhesive wetting period with a linear pressure increase to 0.5N, and a final curing period with a pressure maintained at 1.2N.

9. The intelligent eye-controlled bonding system for orthodontic appliances according to claim 1, characterized in that: The system also includes a biosafety protection unit for covering the surface of the robotic arm with a medical-grade silicone isolation film, and integrating ultraviolet light-emitting diodes (LEDs) for dynamic sterilization at the end.

10. A bonding method of an intelligent eye-controlled bonding system for an orthodontic appliance according to any one of claims 1 to 9, characterized in that: include: S1 uses an infrared camera array combined with a corneal reflection algorithm to capture the doctor's pupil movement trajectory in real time and analyze the two-dimensional coordinates of the doctor's gaze point through a deep learning model; S2, real-time monitoring of the dynamic changes of the physician's head, and correction of the two-dimensional coordinates of the physician's gaze point through homogeneous coordinate transformation to generate the compensated coordinates of the physician's gaze point; S3, through the spatial coordinate system conversion algorithm, the doctor's gaze point compensation coordinates are mapped to the three-dimensional coordinates on the target tooth surface to generate the initial three-dimensional coordinates of the bracket target bonding point; S4, based on the initial three-dimensional coordinates of the bracket target bonding point, the digital dental model obtained by the intraoral scanner and the cone-beam CT data were integrated to construct a three-dimensional visualization interface; S5, projecting the initial three-dimensional coordinates of the bracket target bonding point onto the physical tooth surface in real time using a laser positioning projection device; S6, dynamically tracking the patient's maxillofacial movements. When dynamic changes in the patient's maxillofacial region are detected, the corrected position of the bracket's target bonding point is generated based on the initial three-dimensional coordinates of the bracket's target bonding point through the superposition of rigid compensation and elastic compensation, and is synchronously updated to the three-dimensional visualization interface and laser positioning projection device; S7, based on the corrected position of the bracket target bonding point, an improved RRT*-Connect algorithm is used to generate an initial collision-free path between the robotic arm and the oral soft tissue, and the elastic map dynamic obstacle avoidance and impedance-admittance hybrid control strategy are integrated to optimize the initial collision-free path in real time, generate an optimized motion path, and display the optimized motion path on a three-dimensional visualization interface; The S8 adopts a modular six-axis collaborative architecture, integrates a triple composite drive system and a dual-cavity end effector, and performs bracket bonding operations with sub-millimeter precision in the confined space of the oral cavity based on the optimized motion path and the corrected position of the bracket target bonding point displayed on a three-dimensional visualization interface.

Images